The REPL is like unit tests that don’t stay and I like unit tests that don’t stay because I don’t need them once my high-level ones work, and they can prevent refactoring.

If you want to know how property testing works in reality (as opposed to checking that reverse . reverse = id), Oskar Wickström is writing a series of posts on property testing the ugly parts of your app.

So true. Finding the simple, beautiful essence of a system is another joy: one that can be better for us, our organizations, and our users.

A definition of done at Microsoft: “Live in production, collecting telemetry that examines the hypothesis which motivated the deployment.” Donovan Brown #BlackLivesMatter #deliverAgile Yes! Start with “how will I know this is delivering value”

When we rush development, skip tests and refactoring, we get “Escalating Risk.” Please give up the “technical debt” description; it gives businesspeople a very wrong impression of the tradeoffs. From @Janellekz #deliverAgile

When you’re as happy to be wrong as you are to be right ... 😉

The best summary of the research around Mel Conway that I’ve seen. goodroot.ca/post/2018/10/1…

When I was younger, I was very interested in learning new technologies as a way of becoming a better software developer. These days, I’m more interested in learning how to think better, in general. (thread)

“You spend most of your money on the things that buy you differentiation But you should only spend your time, your knowledge, your wisdom on the things that are your differentiation.” Buy the rest; money is fungible, knowledge is not. #QConNYC

When someone asks you what your leadership style is, the ideal answer is a flexible style. Don't get wedded to one way of making decisions and motivating people. Don't privilege your preferences over what's productive. Effectiveness depends on adaptability. #WednesdayWisdom

I just built a *functioning* React app by describing what I wanted to GPT-3. I'm still in awe.

Wizardry

A lot of the engineers I respect the most are slower now than they were a decade ago. There's a circumspection, a compulsion to get perspective on the problem. In the macro it saves time. In the micro they are sometimes slower than a junior at the same task.

I used to aim for abstraction, so code could be more general. Now I aim for expressiveness, so the more specific the better.

I just bought "Sam Howell Core" by Candy bitski.com/Candy/products…

IRL, science is a lot of hard work finding ways to measure the information we need from the real world. In software, we get to make the world we study tell us what’s going on. If we could inject telemetry into elementary particles, physics would be a different game.

Each code change doesn’t have to be perfect. We need it to be a little better than the code was before.